On May 20, 2026, SpaceX filed its S-1 with the SEC. The filing showed a company that did $18.7 billion in revenue the prior year — up 33% — and still posted a $4.9 billion GAAP net loss, because it is simultaneously building Starship, a mobile phone network, and data centers in orbit. The reported valuation target: roughly $1.75 trillion. That is not a typo. It is, if achieved, one of the largest public listings in history, and it tells you everything about where space investing stands in 2026: no longer a curiosity trade for enthusiasts, but a capital formation story that institutional investors can no longer afford to ignore.

The macro frame keeps getting bigger. The World Economic Forum and McKinsey pegged the global space economy at about $630 billion in 2023, projecting it could reach $1.8 trillion by 2035. The Space Foundation measured roughly $613 billion in 2024, up 7.8% year over year and nearly double the 2014 level. Three-quarters-plus of that activity is commercial — J.P. Morgan's private bank puts the commercial share at roughly 78% — which means the growth is being driven by paying customers, not appropriations. Government spending still matters; it acts as a durable demand floor. But the commercial layer on top is moving faster.

Start with the business that is actually profitable. Starlink did about $11.4 billion in revenue in 2025, up roughly 83% from the year before, and its adjusted EBITDA margin reached an estimated 63% — a number that belongs in the software industry, not in satellite manufacturing. Subscriber counts tell the adoption story: 2.3 million at the end of 2023, 4.6 million at the end of 2024, 9.2 million at the end of 2025, and more than 10 million by February 2026, spread across 160 countries. That is a compounding consumer business with recurring subscription revenue. Everything else in the space portfolio — launch, Starship, the AI compute bet — is being financed, in part, by satellite internet subscribers paying their monthly bills.

The Three Pools Where Capital Is Actually Accumulating

First: connectivity and the data stack around it. Starlink's Q2 2026 connectivity revenue alone ran at roughly $4.3 billion in a single quarter, up 66% year over year. The next-generation V3 satellites — each carrying roughly a terabit of downlink capacity and deployable only on Starship, which reached orbit and deployed 26 of them in September 2026 — underpin plans for gigabit consumer service and a target of about 25 million users by year-end 2026. Meanwhile the July 2025 T-Mobile T-Satellite launch made satellite-to-phone service commercial in the US, and in August 2026 Airtel Africa launched commercial dish-free satellite-to-phone service in the Democratic Republic of Congo — the first of its kind on the continent. The EchoStar spectrum purchase, valued at roughly $17 billion and approved in May 2026, positions the constellation as a direct mobile carrier. The takeaway for investors: the revenue per satellite keeps getting levered upward as the addressable market expands from ~400 million households to ~5 billion handsets.

Second: the picks-and-shovels layer, broadly defined. Space-tech was the second-leading segment within defense-tech venture activity in Q1 2026 — 40 deals totaling roughly $2.7 billion, per J.P. Morgan's private bank. One industry tally put total 2026 space startup funding at a record $20.3 billion, with orbital compute the fastest-growing category. Suppliers of phased-array antennas, laser crosslinks, propulsion, ground terminals, and debris tracking all sit downstream of a simple truth: there are now more than 8,000 active satellites in orbit, launch costs have fallen by more than 90% over the past decade, and the marginal cost of getting hardware up keeps dropping. Every price decline in launch is a price increase in addressable demand for everything that rides the rockets.

Third: defense and sovereign demand, which has become the ballast. Governments of nearly 100 countries have now operated a satellite, per OECD tracking, and national-security budgets for space keep growing across the US, Europe, India, and Japan. SpaceX's defense-oriented Starshield unit is part of this demand, and it matters for investors because government contracts provide the multi-year revenue visibility that venture-backed startups typically lack.

The Skeptic's Case Deserves a Hearing

Here is the uncomfortable counterargument, and it is worth stating plainly: valuation is not value creation, and the distance between the two is where space investors go to lose money. The S-1 itself is candid — the AI segment posted a $6.4 billion operating loss in 2025 and $7.7 billion of capex in a single quarter of 2026. Orbital compute is a thesis, not a business, and independent analysis suggests space-based data centers handle only single-digit percentages of AI workloads even in aggressive projections through 2035. Launch has improved 98% on cost and still operates on thin economics for anyone without reusable scale. Most importantly, the $1.8 trillion forecasts count economic activity enabled by space services — navigation improving logistics, satellite data improving agriculture — not revenue accruing to space companies. Investors get paid from the second, smaller number.

So where is the smart money actually going in 2026? Toward cash-generative connectivity businesses with proven unit economics; toward infrastructure suppliers whose customers have signed contracts rather than visions; and toward defense-adjacent plays where the buyer is a government with a budget line. It is moving away from pure-play speculation on timelines nobody controls. The discipline of this cycle is what the last one lacked.

Space is no longer an asset class you buy for the story. It is an asset class you underwrite like infrastructure — and the investors who treat it that way are the ones who will still be in it when the next cycle starts.

spaceeconomycapital.com is available for acquisition by the company that intends to own the space-investment conversation. Inquiries: info@primedomaincapital.com.